An Investigation of Mothers' Communication Orientations and Patterns

ABSTRACT - This study investigated the efficacy of communication orientations and a typology of family patterns based on communication differences. Previous studies of these orientations and typology have been based on the responses of adolescents. Mothers served as the subjects in this investigation. Results from exploratory and confirmatory factoring as well as preliminary construct validation of the orientations and typology provided some support for their use with samples other than adolescents. Extensions and implications of the findings for consumer socialization are suggested.

Citation:

Les Carlson and Sanford Grossbart (1990) ,"An Investigation of Mothers' Communication Orientations and Patterns", in NA - Advances in Consumer Research Volume 17, eds. Marvin E. Goldberg, Gerald Gorn, and Richard W. Pollay, Provo, UT : Association for Consumer Research, Pages: 804-812.

Advances in Consumer Research Volume 17, 1990      Pages 804-812

AN INVESTIGATION OF MOTHERS' COMMUNICATION ORIENTATIONS AND PATTERNS

Les Carlson, University of Arkansas

Sanford Grossbart, University of Nebraska

Carolyn Tripp, University of Arkansas

ABSTRACT -

This study investigated the efficacy of communication orientations and a typology of family patterns based on communication differences. Previous studies of these orientations and typology have been based on the responses of adolescents. Mothers served as the subjects in this investigation. Results from exploratory and confirmatory factoring as well as preliminary construct validation of the orientations and typology provided some support for their use with samples other than adolescents. Extensions and implications of the findings for consumer socialization are suggested.

INTRODUCTION

Marketers have sought to determine why differences in consumer behavior develop. Consumer socialization provides useful insights in this regard by attempting to identify and clarify those processes by which an individual acquires a consumer identity, complete with differential opinions, knowledge and skills. This study, by focusing on mothers' perspectives, provides additional insights on family communication about consumer learning.

Background

One research stream focuses on the influence role of various communication agents (e.g., parents) in consumer socialization (e.g., Moschis and Moore 1978; Moore and Moschis 1981; Moschis, Moore and Smith 1983; Moschis, Prahasto and Mitchell 1985; Moschis and Mitchell 1985). Since consumer socialization is the process "by which young people acquire skills, knowledge and attitudes relevant to their functioning in the marketplace," (Ward 1980, p. 380) understanding the nature of parent-child communication may provide an explanation for differences in consumer behavior and skills. In this context, Chaffee et al. (1971) and others (McLeod and Chaffee 1972; McLeod and O'Keefe 1972) provide a useful framework which describes family communication orientations and patterns in political socialization. This framework has been adapted to a consumer socialization context (cf., Moschis 1985; Moore and Moschis 1981). A review of the conceptual bases of these orientations and patterns follow.

Consumer Socialization Communication Orientations

These patterns are derived from two uncorrelated family communication dimensions. Socio-oriented parental messages promote deference to parents and monitoring and control of children's consumer learning while concept-oriented parental communications encourage children to develop their own consumer skills and competencies. These dimensions can be arrayed to form a two by two typology depicting four parental communication patterns: Laissez-faire, Protective, Pluralistic, and Consensual.

Laissez-faire parents are neither socio- nor concept-oriented and hence little parent-child communication of either type occurs in these families. Protectives emphasize deference to parents, control and mediation of children's consumer socialization at the expense of building consumer competence and skills in the child. Pluralistics encourage the development of consumer skills without promoting acquiescence to themselves or controlling children's exposure to the marketplace. Finally, Consensuals stress both orientations. Their children are encouraged to take interest in ideas outside the scope of family beliefs and traditions, yet simultaneously are expected to learn and conform to parental consumer behaviors, values and ideas.

These patterns are believed to account for differential outcomes in offspring beyond those attributable to parent-child interactions such as modeling and reinforcement strategies and have been cited as directly or indirectly impacting consumer learning (Moschis 1984; Moschis and Moore 1978; 1983; Moschis, Moore and Smith 1983; Moschis, Prahasto and Mitchell 1985). Direct influences involve acquisition of-consumption related information and formation of beliefs, norms and behaviors while indirect influences include affecting and mediating children's interactions with other - sources of consumer influence, e.g., media (Moschis 1985).

Yet, while these findings provide understanding of the influence of these patterns on consumer learning, this research has been conducted almost exclusively on adolescents' views of family communication. In the only variations to this sampling frame, Moschis and Mitchell (1985) used mother and adolescent pairs; however, only the youngsters responded to communication orientation items. Similarly, Foxman et al. (1989), using mother, father and adolescent triads, found that family communication patterns were an influencing factor on consumption decisions. Again, while data were gathered from a variety of family members, pattern assignment was based on adolescent responses to socio and concept measures.

In summary, while these patterns have been related to consumerlearning and family influence relations they have been investigated from the perspective of adolescents. Thus, only one partner (adolescents) of the communication dyad (i.e., parent-child) in consumer socialization has been studied. Likewise, validity assessment of these orientations and of the typology of family communication patterns has been limited to samples of adolescents. Also, key assumptions across some research are the orthogonality and unidimensionality of the communication orientations which are essential preconditions for assigning individuals to the patterns in the typology. Thus, there appears to be several gaps in our understanding of these orientations and in their application. By focusing on mothers, this study sought to provide additional insights into these orientations and patterns.

METHOD

As part of a larger study, self-administered questionnaires were distributed to mothers via their children in three elementary schools from a variety of socioeconomic areas in a mid-western city. Mothers were asked to answer items with respect to their youngest school aged child to avoid multiple responses from the same family and because of previous research emphasis on adolescents' views of their family communication environments.

In addition to the need to study mothers' perspectives using the socio- and concept-orientations, mothers were selected for these reasons. In consumer socialization terms, mothers are important because of their greater familiarity with the marketplace and its relation to children (Aldous 1974), their roles as mediators of consumer socialization agent influence (Abrams 1984), and their long-term impact on the consumption choices of children (Alsop 1988).

Questionnaires were collected two weeks after distribution and the 46% response rate (n = 499) was similar across schools. Principals considered this rate to be normal compared to other take home items requiring a parental response and it was similar to rates reported previously (cf., Grossbart and Crosby 1984). 451 cases were retained for analyses, incomplete items occurred in a random fashion and there were no differences on dependent variables between respondents who did/did not complete all items.

Communication Pattern Variables

The most recent versions of the concept and socio scales (Moschis, Moore and Smith 1983) were slightly revised to allow mothers, rather than adolescents to serve as respondents. Six concept items reflected maternal emphasis on developing children's competence and skills as consumers (e.g., I tell mychild to decide about things he/she should or shouldn't buy) while five socio items tapped maternal emphasis on monitoring and controlling children's consumptive behavior (e.g., I want to know what my child does with his/her money) (see Table 2). Responses were summed across items in each scale. One socio item (as cited in Moschis, Moore and Smith 1983) was deleted because it was deemed reflective of general parenting orientations rather than specifically oriented toward consumption issues (i.e., Parents say they know what is best for the child and he/she shouldn't question them).

As these measures were to be used to classify mothers per one of the communication patterns a check was performed on the potential impact of responding to these measures in a socially desirable manner. Correlations were calculated between each index and a 19 item version of Crowne and Marlowe's (1964) social desirability scale. Social desirability accounted for less than 1% of the variation in socio-orientation and less than 2% of the variation in concept-orientation. Another correlation analysis investigated the possible confound of mothers couching their responses according to their youngest school aged child. However, neither communication orientation was related to birth order of the child or the number of children in the family.

Indicators of Media Mediation, Monitoring and Control

Moore and Moschis (1981) provided initial validity evidence for socio and concept indices by relating communication patterns (e.g., Protectives) to the adolescent's media use (e.g., amount of tv viewing). In an attempt to be consistent with procedures used by these authors, differences across patterns were examined for mother's reports of children's aggregate tv viewing, parent-child discussions about advertising, viewing tv with children and control of children's tv viewing. These variables appeared to be consequences and/or natural outcomes of fostering consumer competencies vs. controlling children's consumer learning (as tapped by the concept- and socio-orientations). Although differences across patterns were not hypothesized, certain relations were intuitively appealing. For example, Protectives and Consensuals ought to be higher on variables reflecting maternal control tendencies, i.e., amount and restriction of children's tv viewing, as they emphasize socio-oriented messages. Likewise, Pluralistics and Consensuals should be higher on tendencies to foster the development of the child's consumer skills and knowledge, i.e., parent-child discussions about advertising and viewing tv with children since they stress concept-oriented messages. Laissez-faires might be low on all tendencies as they refrain from either type of message.

Amount of children's tv viewing was indicated by number of hours a child viewed tv for a typical weekday and weekend. Mothers also noted the extent they discussed tv, magazine, newspaper and radio advertising with children. Coviewing tapped the frequency mothers watched tv with children on weekdays and weekends and the importance mothers placed on this activity. Finally, mothers cited extent of restrictions placed on children's tv viewing in terms of programming, times and amounts (see Table 1).

Analyses

The analytical procedure was composed of three parts. First, alpha and beta reliability analyses and exploratory factoring were used to examine item patterns and indicate whether item responses from mothers to the socio and concept scale items were similar to those obtained from adolescents in prior studies. Next, since the four communication patterns are based on a theoretical orthogonal array of the two communication dimensions, confirmatory factor analysis was used as a test of this structure. Support for this array (versus competing depictions) would provide evidence for using it when investigating the impact of communication on consumer socialization. Finally, following previous investigations of communication patterns (e.g., Moschis and Moore 1979), median splits on the socio- and concept-orientation raw scores were used to place mothers into one of the four cells of the communication typology. Means on the various media variables were compared via one-way ANOVAs across communication patterns as an indicator of construct validity (cf., Moore and Moschis 1981).

TABLE 1

COMMUNICATION PATTERN AND MEDIA VARIABLES

RESULTS

Alpha reliabilities for the concept and socio scales using mothers (see Table 1) were very similar to those reported by Moschis and Mitchell (1985; concept alpha = .72; socio alpha = .51) for adolescent respondents. Initial exploratory factoring suggested a four factor solution accounting for 59.3% of the variation. Both concept and socio communication items loaded on two factors, suggesting these orientations may be multidimensional. The socio beta of .49 also raises questions about its unidimensionality (Table 1). Loading patterns were relatively clean, e.g., socio items loaded primarily with other socio items as opposed to cross loading with concept items (see Table 2). Confirmatory factor analyses (as discussed below) suggested a four factor oblique solution, thus, rotated oblique results from exploratory factoring are presented in Table 2.

Confirmatory factoring via LISREL VI was used to examine several competing conceptual views. First, as noted, it could provide an indication of the appropriateness of the orthogonal array. Second, it might be argued that these orientations are not orthogonal dimensions but are merely end points or locations on a single continuum. That is, only one underlying consumer socialization communication dimension exists and mothers merely vary between controlling and fostering children's consumer learning. Finally, perhaps these items do not represent any underlying structure. Confirmatory factoring allows for comparing hypothesized models to each other and to a "null" model where complete independence is assumed between all indicators.

TABLE 2

OBLIQUE FACTOR RESULTS

TABLE 3

CHI-SQUARE AND GOODNESS OF FIT VALUES FOR ALTERNATIVE MODELS

Fit indices (i.e., Adjusted Goodness of Fit Index - AGFI) and a ratio using the null model as a reference (Bentler and Bonett 1980; Shimp and Sharma 1983) were used to compare models. Exploratory factoring (i.e., eigenvalues greater than one) suggested additional models to the null and two factor orthogonal possibilities. The results of these comparisons are presented in Table 3. The null model, postulating that each of the 11 items from the socio and concept scales are unitary constructs (i.e., that there is not a more parsimonious explanation for the data structure), served as a comparison for subsequent models. The one factor model represented the unitary communication construct possibility. The two factor orthogonal model was suggested by previous theory and represents the underlying structure of the four communication patterns. The four factor models were suggested by exploratory results. Both oblique solutions (models 4 and 6) were attempts to improve fit by allowing for intercorrelation (oblique rotation) between the dimensions.

All models, regardless of rotational methods or number of dimensions, were improvements over the null. In every case, both fit indices indicated improvement in fit between the data and a hypothesized model compared to the null. Allowing for intercorrelation between factors did not significantly improve the two factor model (model 4, Table 3) but did impact the four factor model (model 6, Table 3).

Finally, construct validity of the communication framework was assessed using a procedure similar to Moore and Moschis (1981). Four one-way ANOVAs were performed using communication pattern as an independent variable. Amount of tv viewing, discussions about advertising, coviewing of tv between mothers and children andmother's control of children's tv viewing were the dependent variables. These results are depicted in Table 4. No differences between communication patterns were observed for amount of children's tv viewing. However, Laissez-faires were lower than all groups except Protectives on discussions about advertising and were lower than Pluralistics on coviewing. Laissez-faires were also lower than Pluralistics and Consensuals on controlling viewing. Pluralistics were higher than Protectives on discussing advertising while Protectives and Laissez-faires were higher than Pluralistics on controlling viewing.

DISCUSSION

The objectives of this study were three fold, First, we attempted to explore the socio- and concept-communication orientations from the perspective of mothers due to their significance in the consumer socialization process. Second, we investigated the efficacy of the four cell communication pattern typology (as proposed by Moschis and others), plus other depictions of parent-child communication. Finally, we studied the relations between communication patterns and mother's mediation, monitoring and control of children's media use as a preliminary step toward validation of the socio and concept communication scales. The following discussion considers each of these issues.

TABLE 4

MOTHER'S MEDIATION, MONITORING AND CONTROL OF CHILDREN'S MEDIA USE

Maternal Communication Patterns

As noted, exploratory and confirmatory factoring suggested a four dimensional array. This model (model 6, Table 3) reflects splits in each scale, suggesting that mothers' communication orientations may be multidimensional (see Table 2). For example, one concept dimension reflects mothers' tendencies to instruct children to make decisions about purchases irrespective of the opinions of others (e.g., Item 1, I tell my child to decide about things he/she should or shouldn't buy). The other concept dimension reflects mothers' tendencies to solicit children's opinions regarding purchasing/consumption decisions that may or may not be related to the child (e.g., Item 7, I ask my child to help me buy thingS for the family).

The socio-orientation appears to be composed of maternal tendencies toward control of children's consumption (e.g., Item 4, I tell my child what things he/she should or shouldn't buy). Mothers also appear to be exerting control via cold indifference toward their children's consumption activities (e.g., Item 10, I tell my child he/she shouldn't ask questions about thingschildren do not usually buy and Item 11, I want to know what my child does with his/her money -- negative loading). The child may be expected to know that certain consumption activities are simply not discussed because the limitations on his/her behavior have been previously established and are nonnegotiable.

These splits may be due to differences in how mothers vs adolescents perceive communication orientations in a household, i.e., mothers may have a broader perspective on nuances in familial communication tendencies. As such, maternal orientation multidimensionality may suggest finer and more numerous communication pattern delineations than the four typically derived from adolescent responses. Specifically, this might be manifested in communication patterns defined by variations on the four communication orientations indicated in Table 2. Within construct multidimensionality may also suggest the need for scale revisions (e.g., addition of other items) for respondents other than adolescents.

Confirmatory Factoring

As expected, the null model depicted the worst overall fit. The best fit was the four factor oblique solution with the constraints on model estimation noted in Table 3 (superscriptc). Exploratory and confirmatory factor analyses all indicated the within construct multidimensionality of both orientations and multidimensionality of the socio-orientation is also suggested by its beta reliability of .49 (Table 1). Confirmatory factoring also suggested these orientations are not merely locations on a unitary communication dimension as was depicted in the single factor model. Unfortunately, the two factor orthogonal array which was based on prior theory and used in past research to categorize adolescents' perceptions of their family communication environments into one of the four communication patterns, does appear to be "improvable". While this finding does not ease classification of mothers into communication patterns it may suggest mothers perceive communication about consumption issues differently than their offspring (although this study dealt with mothers of elementary school children).

Yet, even though these orientations appear to be multidimensional there are empirical and conceptual reasons to not discard the two dimensional orthogonal array. First, the exploratory factoring scree plot suggested a two factor orthogonal solution. Second, although the four factor oblique solution delineates a 56 point x2 improvement compared to the two factor orthogonal array, the improvement in the AGFI and Bentler and Bonett (1980) fit indices is not substantial (see Table 3, models 3 and 6). Third, the correlations between factors for the four factor oblique solution are quite low (see Table 2). The highest correlation (i.e.,between concept factors) indicates that less than 7% of the variation in one concept factor is accounted for by the other even though these factors were allowed to interrelate. Hence, although several analyses suggested within construct multidimensionality with intra-construct relationships, socio and concept dimensions are only marginally related. This supports the basic independence of the concept and socio tendencies but does not necessarily provide additional evidence for a two dimensional structure.

This evidence suggests a research dilemma, i.e., using multidimensional socio and concept scales which may be a more accurate depiction of how mothers perceive these constructs vs a concomitant loss of subject classification ease and model interpretability. Given these liabilities should researchers refrain from classifying mothers per the typical four patterns? To provide some guidance, as noted, we conducted additional tests of the typology's efficacy by investigating the relations between communication patterns and mother's mediation, monitoring and control of children's media use. These tests provide preliminary indications on the construct validity of these orientations and the typology.

Media Mediation, Monitoring and Control by Communication Pattern

Surprisingly, there were no differences between communication patterns on mother's indications of amount of children's tv viewing (see Table 4). This is contrary to previous findings (e.g., Moore and Moschis 1981) and may be a reflection of using mothers and querying them on the viewing habits of their youngest elementary age child. The lack of significant differences may also be a result of the within construct multidimensionality issue cited previously. Our results indicate children's tv viewing does not vary by mother's communication type. What is not known is whether mothers report differences in adolescent viewing across communication patterns.

However, differences across patterns were found for other media use variables. Note that relations between these variables and patterns are not simple reflections of maternal response consistencies due to assumed similarities between the independent and dependent measures. For example, the discussions about advertising index asked mothers for the frequency they talk to children about the content of tv, magazine, newspaper and radio advertising. This variable is conceptually distinct from either the concept or socio indices (i.e., I tell my child to decide about things he/she should or shouldn't buy and I tell my child what things he/she should or shouldn't buy).

Laissez-faires, who by definition, are lowest on bothcommunication orientations are also lower on discussing advertising with children than all groups except Protectives. Very little parent-child communication occurs in Laissez-faire families (Moore and Moschis 1981). Hence, it is not surprising that these mothers report few interactions with children about advertising. Likewise, this general lack of communication is also manifested in controlling tv viewing less than all other groups except Pluralistics though Laissez-faires coview less than Pluralistics.

Pluralistics might coview with children because this activity provides a springboard for parent-child discussions about advertising. Pluralistics foster consumer learning without emphasizing monitoring and control of consumption behavior. Thus, it's reasonable that Pluralistics discuss ads more than Protectives and Laissez-faires who both deemphasize concept messages.

Protectives emphasize control and monitoring of consumer learning via socio-oriented messages at the expense of building consumer skills. Not surprising then, Protectives are higher than all groups except for Consensuals on control of tv viewing but control apparently does not translate into differences on children's tv viewing. Both Protectives and Consensuals emphasize socio-oriented messages and hence might be expected to be similar on control of tv viewing.

Consensuals were higher than Laissez-faires on discussing ads and control of tv viewing. This is reasonable since Consensuals emphasize, while Laissez-faires deemphasize, the communication tendencies that might be related to these concepts, i.e., concept- and socio-oriented messages, respectively. Consensuals were higher on coviewing than all groups and the lack of significant differences may be a function of differential sample sizes (see Table 4). Like Pluralistics, they too may be using coviewing as a stimulus for discussions about ads.

The lack of additional media related differences may be due to the reasons previously cited for finding no differences on amount of children's tv viewing (e.g., within construct multidimensionality). Still, even with these limitations, the differences that were found could be reasonably expected given concept and socio variations between the four communication patterns. Thus, researchers may have some confidence in the construct validity of these orientations and in the Practice of assigning mothers to one of the patterns.

SUMMARY, EXTENSIONS AND CONCLUSIONS

Since past research has often used adolescent respondents thisstudy attempted to extend communication pattern research by sampling mothers with elementary age children. We revised communication orientation items to facilitate their use by mothers rather than adolescents. We also investigated the efficacy of the pattern typology which can be constructed from these communication orientations and several alternative arrays. Finally, because of possible multidimensionality of concept and socio tendencies, we have provided other potential indicators of their construct validity (but with a sampling frame different from previous research).

This research may be extended in several ways. First, while these orientations and patterns have now been studied from the perspective of adolescents and mothers, their applicability to other family members, i.e., fathers, also should be investigated. Second, corroboration between multiple respondents within the same family unit would give additional insight into the true nature of communication within a household. This would help to answer a number of important questions. Do children confirm the type of communication about consumption that is reported by their parents? Do female and male heads of households note similar use of each communication orientation? If disagreement exists, how does it affect the consumer socialization process? Third, further refinement of these scales may be needed in order to address the within pattern multidimensionality issue. Additions, deletions and/or reexaminations of items may be necessary to obtain un dimensional constructs. Construct validation of these orientations might be enhanced with measure revisions.

Communication orientations and patterns are a rich area for additional conceptualization and research. The findings cited here are extensions to previous communication and socialization research. For example, our results suggest differences in mother's mediation, monitoring and control of children's media use may be a partial function of general communication differences. The orientations and framework investigated here and in other communication pattern studies should continue to provide new insights into the role of parents as socialization agents in children's consumer learning.

REFERENCES

Abrams, Bill (1984), "TV Ads, Shows Struggle to Replace Bygone Images of Today's Mothers," The Wall Street Journal, 204 (October 5), 27.

Aldous, Joan (1974), "Commentaries on Ward, 'Consumer Socialization'," Journal of Consumer Research, 1 (September), 15-16.

Alsop, Ronald (1988), "Mom Leaves Her Mark in Loyalty to Products," The Wall Street Journal, (January 19), 33.

Bentler, P.M. and Douglas G. Bonett (1980), "Significance Tests and Goodness of Fit in the Analysis of Covariance Structures," Psychological Bulletin, 88 (no. 3), 588-606.

Chaffee, Steven H., J. M. McLeod, and C. K. Atkin (1971), "Parental Influences on Adolescent Media Use," American Behavioral Scientist, 14, 323340.

Crosby, Lawrence A. and Sanford L. Grossbart (1984), "Parental Style Tendencies and Concern About Children's Advertising," in Current Issues in Research in Advertising, eds. James H. Leigh and Claude R. Martin, Jr., Ann Arbor, MI: Division of Research, Graduate School of Business Administration, 43-63.

Crowne, Douglas P. and David Marlowe (1964), The Approval Motive: Studies in Evaluative Independence, New York: John Wiley.

Foxman, Ellen R., Patriya S. Tansuhaj and Karin M. Ekstrom (1989), "Adolescents' Influence in Family Decisions,A Socialization Perspective," Journal of Business Research, 18, 159-172.

Grossbart, Sanford L. and Lawrence A. Crosby (1984), "Understanding the Bases of Parental Concern and Reaction to Children's Food Advertising," Journal of Marketing, 48 (Summer), 79-92.

McLeod, Jack M. and Steven H. Chaffee (1972), 'The Construction of Social Reality," The Social Influence Process, J.T. Tiedeschi, ed., Chicago, 50-99.

McLeod, Jack M. and Garrett O'Keefe, Jr. (1972), 'The Socialization Perspective and Communication Behavior," in G. Kline and P. Tichenor, Current Perspectives in Mass Communication Research, Beverly Hills, CA, Sage Publications.

Moore, Roy L. and George P. Moschis (1981), "The Role of Family Communication in Consumer Learning," Journal of Communication, 31 (Autumn), 42-51.

Moschis, George P. (1984), "A Longitudinal Study of Consumer Socialization," in Proceedings of the American Marketing Association Conference, eds., Paul F. Anderson and Michael J. Ryan, Chicago: American Marketing Association, 189-192.

Moschis, George P. (1985), "The Role of Family Communication in Consumer Socialization of Children and Adolescents," Journal of Consumer Research, 11 (March), 898-913.

Moschis, George P. and Linda G. Mitchell (1985), "Television Advertising and Interpersonal Influences on Teenagers' Participation in Family Consumer Decisions," in Advances in Consumer Research, Vol. 13, ed. Richard Lutz, Provo, UT: Association for Consumer Research, 181-191.

Moschis, George P. and Roy L. Moore (1978), "Family Communication & Consumer Socialization," in Advances in Consumer Research, Vol. 6, ed. William L. WiLkie, Ann Arbor, MI: Association for Consumer Research, 359-363.

Moschis, George P. and Roy L. Moore (1979), "Decision Making Among the Young: A Socialization Perspective," Journal of Consumer Research, 6 (September), 101-112.

Moschis, George P. and Roy L. Moore (1983), "A Longitudinal Study of the Development of Purchasing Patterns," in Proceedings of the American Marketing Association Conference, eds., Patrick E. Murphy, Gene R. Laczniak, Paul R. Anderson, Russell W. Belk, O.C. Ferrell, Robert F. Lusch, Terence A. Shimp and Charles B. Weinberg, Chicago: American Marketing Association, 114-117.

Moschis, George P., Roy L. Moore, and Ruth B. Smith (1983), 'The Impact of Family Communication on Adolescent Consumer Socialization," in Advances in Consumer Research, Vol. 11, ed. Thomas C. Kinnear, Provo, UT: Association for Consumer Research, 314-319.

Moschis, George P., Andjali E. Prahasto, and Linda G. Mitchell (1985), "Family Communication Influences on the Development of Consumer Behavior: Some Additional Findings," in Advances in Consumer Research, Vol. 13, ed. Richard Lutz, Provo, UT: Association for Consumer Research, 365-369.

Shimp, Terence A. and Subhash Sharma (1983), 'The Dimensionality of Involvement: A Test of the Automobile Involvement Scale," in William R. Darden, Kent B. Monroe, and William R. Dillon, AMA Winter Educators' Conference: Research Methods and Causal Modeling in Marketing, Chicago: American Marketing Association, 5861.

Ward, Scott (1980), "Consumer Socialization," in Perspectives in Consumer Behavior, Third Edition, Harold H. Kassarjian and Thomas S. Robertson, eds., Glenville, IL: Scott, Foresman and Company.

Ward, Scott, Daniel B. Wackman, and Ellen Wartella (1977), How Children Learn to Buy, Beverly Hills, CA: Sage.

----------------------------------------